kiranvad / pymecsim Goto Github PK

This is simply because the subprocess call can't locate the MECSIM binary file

Currently the api involves initiating a mechanism with seperate lists of species and reactions. Species can be obtained from the mechanism. So we change the api to the following:

A = Specie('A', C0=1e-6)
B = Specie('B')
C = Specie('C')
D = Specie('D')

R1 = ChargeTransfer([(A,1),('e',1)],[(B,1)],E0=0.0, ks=1.0e1)
R2 = ChemicalReaction([(B,1),(C,1)],[(A,1),(D,1)], kf=1e8, kb=1e-5)

A reaction now takes tuple of the form (specie, coefficient) instead of dictionary with {'specie.name':coefficient}.

A mechanism can now be initiated by just passing the list of reactions as follows:

rxn = [R1, R2]
mech = Mechanism(rxn)

Also few performances issues needs to be addressed:

1. Species order to generate input file should be solution species followed surface reaction species.
2. Species should also follow the order : species on the left side of equations should come first in the

   pyMECSim/pymecsim/input.py

   Line 214 in 2ea3be3

   self.species = species

3. input files that change for every simulation should be untracked and removed from repository.

The way the package is set up works only with Linux (this is mainly a drawback from the MECSim). Make this explicit in the installation.

• Add matplotlib as a dependency
• Fix sim.solve() in the readme example

Hi, thanks for creating this python wrapper! I tried installing via pip and also via the current git repo directly but am unable to replicate the notebook demo codes. I'm using a linux subsystem within windows to install your package. The error is as such:

How can this be resolved? Thanks!